Lipopolysaccharide binding protein in a surgical intensive care unit: a marker of sepsis?

OBJECTIVES: We investigated the time course of lipopolysaccharide binding protein (LBP) plasma concentrations in patients in the surgical intensive care unit (ICU), their value in discriminating sepsis from systemic inflammatory response syndrome, and their association with severity of sepsis and outcome in these patients compared with interleukin (IL)-6, C-reactive protein, and procalcitonin.

DESIGN: Prospective, observational, cohort study.

SETTING: Academic ICU.

PATIENTS: All 327 consecutively admitted patients.

MEASUREMENTS AND MAIN RESULTS: Serum LBP concentrations were higher in patients who had severe sepsis/septic shock on ICU admission than in patients who never had sepsis (20.5 [8.1-38.8] vs. 14.2 [7.7-22.2] microg/mL, p < .05) but were similar in patients with sepsis without organ failure and those who never had sepsis. After 3 days, LBP levels were similar in all groups. In a receiver operating characteristic curve analysis, LBP concentrations moderately discriminated sepsis from systemic inflammatory response syndrome (area under curve [AUC] = .66) and severe sepsis from sepsis without organ failure (AUC = .71). IL-6 had the highest AUC in discriminating sepsis from other conditions (AUC = .76) and procalcitonin had the highest AUC for discrimination of severe sepsis from sepsis (AUC = .86). LBP concentrations on admission and during the first week were similar in patients with gram-positive and those with gram-negative infections (15.9 [11-26.7] and 37.2 [25.1-62.4] vs. 16.3 [5.3-31.6] and 31.6 [13.4], microg/mL, p > .2). LBP concentrations on admission were similar in nonsurvivors and survivors and did not discriminate ICU mortality. However, the maximum LBP concentration during the first 3 days in the ICU discriminated moderately between survivors and nonsurvivors.

CONCLUSIONS: In the surgical ICU, LBP moderately discriminated patients without infection from patients with severe sepsis but not from patients with sepsis without organ dysfunction. LBP concentrations did not distinguish between gram-positive and gram-negative infections. The correlation of LBP concentrations with disease severity and outcome is weak compared with other markers and its use as a biomarker is not warranted in this patient population.